thermal: new pandas doesn't allow you to use a DataFrame in a boolean context

Test if the data_frame is None instead, which is basically what you
want.
diff --git a/cr2/thermal.py b/cr2/thermal.py
index 75f24dd..5d9be89 100644
--- a/cr2/thermal.py
+++ b/cr2/thermal.py
@@ -86,7 +86,7 @@
 
     def get_data_frame(self):
         """Return a pandas data frame for the run"""
-        if self.data_frame:
+        if self.data_frame is None:
             return self.data_frame
 
         if not self.data_csv: